1. Field of the Invention
The present invention relates to a digital watermarking device, a digital watermark insertion method and a digital watermark detection method.
2. Description of the Related Art
Digital information can be easily reproduced, and is often copied illegally. Various digital watermarking technologies to insert digital watermark information into the digital information have been proposed so that the copyright holder can claim his/her copyright to the person who copied the information illegally. Devices and methods for digital watermark insertion into the digital information include, for example, the devices and the methods described in the Japanese Patent Application Laid-open (Kokai) No. 2000-59610 and No. 2000-59730.
FIG. 5 is a block diagram showing an configuration example of the device described in the Japanese Patent Application Laid-open (Kokai) No. 2000-59610. In this device, an image input section 540 fetches the image data as the original image. A square division section 550 divides the original image into squares. An information input section 560 inputs the predetermined information to be embedded in the image data and an encoding section 520 embeds such information as the digital watermark information in the image data. Here, the encoding section 520 embeds the digital watermark information to every square (every block) divided. A control section 500 makes various controls. According to the device shown in FIG. 5, the digital watermark information is embedded to every piece of divided squares after division of the image data into squares, and the digital watermark information can be detected even when a part of the image data is cut off.
FIG. 6 shows an image stream to record the digital watermark information by the digital watermark information recording method as described in the Japanese Patent Application Laid-open (Kokai) No. 2000-59730. According to this recording method, the reference information to detect the digital watermark information is recorded in every frame containing the time information in the image stream. Here, the reference information is recorded so that it is contained in the encoded information obtained by encoding of the image. The position to record the reference information is uniquely specified by the time information. By this method, the reference information recording position varies frame by frame and the digital watermark information can be taken out by detecting the reference signal even when a part of the image information is masked.
Digital watermark information to be inserted into the image data includes, for example, the information about the owner of the image data and the information about whether or not to permit reproduction of the image data. When many types of digital watermark information are inserted to the image data, it will be difficult for a third party without authority for data utilization to analyze the digital watermark information. Therefore, to prevent analysis by a third party, it is preferable to insert as many types of information as possible for the digital watermark information.
Some types of inserted digital watermark information may deteriorate the image data. By inserting several types of digital watermark information, deterioration extent of the image data by a certain type of digital watermark information can be reduced. From this viewpoint, too, it is preferable to insert several types of information as the digital watermark information.
It is also preferable that the number of types of the digital watermark information to be inserted is highly flexible. Note that, however, the scale of the device for insertion and detection of the digital watermark is preferably small.
According to the device as described in the Japanese Patent Application Laid-open No. 2000-59610, an image is divided into sections and digital watermark information is assigned to each section so that increased types of digital watermark information are inserted. However, because the digital watermark is embedded into the divided images, sufficient amount of digital watermark information cannot be embedded when the original image is not large enough over a certain size. In addition, when a divided section is lacking, the digital watermark information embedded to that section may not be obtained properly. The digital watermark information recording method as described in the Japanese Patent Application Laid-open No. 2000-59730 enables the digital watermark information to be taken out even when a part of the image information is masked, but it does not disclose a method to increase the types of digital watermark information to be recorded.